70 
R. Ruggles Gates. 
offspring a large number of a new form has been called by Bartlett 
(1915) mass mutation. In such a case it is obvious that the original 
change or pre-mutation, an invisible change in a germ cell, must 
have occurred in a plant at least two generations earlier than that in 
which the mass mutation exhibits itself. The 1 or 2 per cent of 
mutants arising from other gigas plants are the result of the chance 
meeting of two germ cells both of which have mutated so as to carry 
the dwarf instead of the tall condition. This interpretation is clearly 
substantiated by the fact that CE. gigas nanella when crossed with 
(E. gigas behaves as a simple Mendelian recessive. It has 28 
chromosomes, like the parent form. 
Thus it would appear that many recessive Mendelian characters 
originate either as single individuals from the union of two 
mutated germ cells, or with a frequency approaching 25% in the 
offspring of individuals which have in turn arisen as a heterozygous 
combination of a mutated and a non-mutated germ cell. 
In races or species where self-fertilization is the rule, a viable 
recessive mutation is very likely to “ come out ” in two or three 
generations, but in many animals, including man, where a large 
amount of inter-crossing of strains takes place, recessive mutations 
may accumulate over a considerable period, and some of them may 
then come out as soon as in-breeding begins. Thus in the case of 
cross-breeding organisms, it is impossible to say, when a recessive 
Mendelian mutation appears, how long it may have been carried in 
the germ plasm in a heterozygous condition without getting a chance 
to express itself. The premutation may have taken place many 
generations earlier. 
As one more case of a presumptive Mendelian mutation in 
CEnothera we may consider CE. brevistylis. Although it has never 
actually occurred as a mutation in controlled cultures, yet the fact 
that it was found by de Vries growing wild with CE. lamarckiana 
although it ripens practically no seeds, indicates that it must have 
been derived from lamarckiana , with which its pollen crosses freely. 
In this way it maintained itself for seventeen years. When seeds 
can be obtained it is shown to breed true, and the chromosomes 
number 14. De Vries (1913) found it to behave as a simple 
Mendelian recessive, and Davis (1918) confirms this from a more 
extensive series of crosses. He finds that the reciprocal crosses 
with lamarckiana are uniform and indistinguishable, the characters 
of lamarckiana being strongly dominant. But measurements show 
that in heterozygous individuals brevistylis has an influence in 
